Abstract
An improvement of the ordinary least squares estimator of the regression vector in a univariate linear model is obtained. This is done by combining observed data with prior information and other additional information. Important features include admissibility in the unrestricted and restricted cases.